Filters








12,257 Hits in 6.5 sec

Transfer Learning from Deep Features for Remote Sensing and Poverty Mapping [article]

Michael Xie, Neal Jean, Marshall Burke, David Lobell, Stefano Ermon
2016 arXiv   pre-print
We demonstrate that these learned features are highly informative for poverty mapping, even approaching the predictive performance of survey data collected in the field.  ...  We train a fully convolutional CNN model to predict nighttime lights from daytime imagery, simultaneously learning features that are useful for poverty prediction.  ...  We would also like to thank NVIDIA Corporation for their contribution to this project through an NVIDIA Academic Hardware Grant.  ... 
arXiv:1510.00098v2 fatcat:7vj33rshxjdatidqwcouasvt2m

Remote Sensing of Urban Poverty and Gentrification

Li Lin, Liping Di, Chen Zhang, Liying Guo, Yahui Di
2021 Remote Sensing  
This paper reviews, summarizes, and evaluates state-of-the-art approaches for identifying and mapping urban poverty and gentrification with remote sensing, GIS, and machine learning techniques.  ...  , and remote sensing can play a key role in this.  ...  Acknowledgments: The authors would like to thank Sarah Di of Carnegie Mellon University for editing and proofreading the manuscript.  ... 
doi:10.3390/rs13204022 fatcat:chdq4j2jizgmrgyxho2z7bmjsa

State-of-the-art and gaps for deep learning on limited training data in remote sensing [article]

John E. Ball, Derek T. Anderson, Pan Wei
2018 arXiv   pre-print
Deep learning usually requires big data, with respect to both volume and variety. However, most remote sensing applications only have limited training data, of which a small subset is labeled.  ...  The first topic is transfer learning, in which some aspects of one domain, e.g., features, are transferred to another domain.  ...  The transfer learning occurred from the ImageNet domain to the domain of night-time lights, and then the night-time lights features were transferred to a third domain, poverty mapping.  ... 
arXiv:1807.11573v1 fatcat:q6vtrod6nvgtrihjafo25iz3wi

Satellite-based mapping of urban poverty with transfer learned slum morphologies

Thomas Stark, Michael Wurm, Xiaoxiang Zhu, Hannes Josef Taubenbock
2020 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
Large-scale slum mapping remains a challenge, fuzzy feature spaces between formal and informal settlements, significant imbalance of slum occurrences opposed to formal settlements, and various categories  ...  Transfer learning can help to improve segmentation results when learning on a variety of slum morphologies, with high F1 scores of up to 89%.  ...  ACKNOWLEDGMENT This research was funded by the European Research Council (ERC) under the European Union Horizon 2020 research and innovation program, grant number 714087-So2Sat.  ... 
doi:10.1109/jstars.2020.3018862 fatcat:4vhtzuxxtfhkzhhqk57ifmftuy

Predicting Poverty Index using Deep Learning on Remote Sensing and Household Data

2019 International journal of recent technology and engineering  
Its mean is 21.09, median 21 and a standard deviation is 1.36. Proposed deep learning inspired model estimates wealth-score for 28393 clusters with an r value i.e.  ...  Another challenge is availability of sufficient amount of data which is solved using transfer learning in Convolutional Neural Network (CNN).  ...  Transfer learning approach is used to overcome Predicting Poverty Index using Deep Learning on Remote Sensing and Household Data Parth Agarwal, Nandishwar Garg, Pratibha Singh the deficiency of data points  ... 
doi:10.35940/ijrte.c3918.098319 fatcat:vo4mr4qgdvcs3du5awdiozorya

MAPPING POVERTY IN THE PHILIPPINES USING MACHINE LEARNING, SATELLITE IMAGERY, AND CROWD-SOURCED GEOSPATIAL INFORMATION

I. Tingzon, A. Orden, K. T. Go, S. Sy, V. Sekara, I. Weber, M. Fatehkia, M. García-Herranz, D. Kim
2019 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
Mapping the distribution of poverty in developing countries is essential for humanitarian organizations and policymakers to formulate targeted programs and aid.  ...  We also propose an alternative, cost-effective approach that leverages a combination of volunteered geographic information from OpenStreetMap and nighttime lights satellite imagery for estimating socioeconomic  ...  ACKNOWLEDGEMENTS We would like to thank Neal Jean, Pia Faustino, Ram De Guzman, Lester James Miranda, and Priscilla Moraes for the insightful discussions and valuable guidance.  ... 
doi:10.5194/isprs-archives-xlii-4-w19-425-2019 fatcat:nufg54g6azgytlq4pw7o3vd7aa

Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks

Michael Wurm, Thomas Stark, Xiao Xiang Zhu, Matthias Weigand, Hannes Taubenböck
2019 ISPRS journal of photogrammetry and remote sensing (Print)  
In the past, remote sensing has proven to be an extremely valuable and effective tool for mapping slums.  ...  Recent advances in deep learning allow for transferring trained fully convolutional networks (FCN) from one data set to another.  ...  Transferring deep features between various remote sensing data sets In slum mapping, in particular approaches using remotely sensed data from satellite images with varying characteristics were used extensively  ... 
doi:10.1016/j.isprsjprs.2019.02.006 fatcat:ejmjor2t7vf3lcmiqurxmedtd4

Evaluation of deep-learning methods to understand the prediction of socio-economic indicators from remote sensing imagery

Jeaneth MACHICAO, Robin JARRY, Danton Ferreira VELLENICH, Jean Pierre OMETTO, Katia FERRAZ, Nadya DEPS, Miguel Suarez Xavier PENTEADO, Shelley STALL, Alison SPECHT, Laurence MABILE, Marc CHAUMONT, Pedro CORRÊA (+1 others)
2020 Zenodo  
learning, so that each input image would correspond to a feature vector and could be annotated with a socioeconomic indicator; and (3) use a simpler regression model to predict poverty measures from the  ...  , aiming to intensively learn the relationship between the input and their images annotations (intermediate outputs); (2) extract a feature vector from the CNN output which will be used as the transferability  ...  Label A Task A Adapted from: https://www.topbots.com/transfer-learning-in-nlp/ Input B Predictions B Label B Task B DL knowledge transfer Learnable (feature extractor) Input A Predictions  ... 
doi:10.5281/zenodo.4280070 fatcat:44zsptccvjfojpopkvuheo6ery

Studying the Effect of Activation Function on Classification Accuracy Using Deep Artificial Neural Networks

Serwa A
2017 Journal of Remote Sensing & GIS  
., Michael Xie [3] , who studied the transfer learning from deep features for remote sensing and poverty mapping.  ...  Both sigmoid and bipolar sigmoid AFs are recommended to be used in classification of remote sensing landcover features.  ... 
doi:10.4172/2469-4134.1000203 fatcat:gpkboch6grh4taq7efbnlw7hyu

2020 Index IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Vol. 13

2020 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
., +, JSTARS 2020 5746-5758 Satellite-Based Mapping of Urban Poverty With Transfer-Learned Slum Morphologies.  ...  ., +, JSTARS 2020 5466-5479 Satellite-Based Mapping of Urban Poverty With Transfer-Learned Slum Morphologies.  ...  A New Deep-Learning-Based Approach for Earthquake-Triggered Landslide Detection From Single-Temporal RapidEye Satellite Imagery. Yi, Y., +, JSTARS 2020  ... 
doi:10.1109/jstars.2021.3050695 fatcat:ycd5qt66xrgqfewcr6ygsqcl2y

Tile2Vec: Unsupervised Representation Learning for Spatially Distributed Data

Neal Jean, Sherrie Wang, Anshul Samar, George Azzari, David Lobell, Stefano Ermon
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We demonstrate empirically that Tile2Vec learns semantically meaningful representations for both image and non-image datasets.  ...  To fill this gap, we introduce Tile2Vec, an unsupervised representation learning algorithm that extends the distributional hypothesis from natural language — words appearing in similar contexts tend to  ...  Acknowledgements This research was supported by NSF (#1651565, #1522054, #1733686), ONR, Sony, and FLI.  ... 
doi:10.1609/aaai.v33i01.33013967 fatcat:j6wopditnnf6xdcszcjzu64cje

Nightlight as a Proxy of Economic Indicators: Fine-Grained GDP Inference Around Mainland China via Attention-Augmented CNN from Daytime Satellite Imagery

Haoyu Liu, Xianwen He, Yanbing Bai, Xing Liu, Yilin Wu, Yanyun Zhao, Hanfang Yang
2021 Remote Sensing  
To date, a series of studies have been conducted to generate fine-grained maps of socioeconomic indicators from the easily accessed remote sensing data and achieved satisfactory results.  ...  To the best of our knowledge, this is the first time that county-level GDP values around the Chinese Mainland have been estimated from both daytime and nighttime remote sensing data relying on attention-augmented  ...  Detection of Economic-Related Visual Patterns from Daytime Satellite Imagery via Deep Learning Remote sensing data are valuable for economic studies because they provide access to information hard to obtain  ... 
doi:10.3390/rs13112067 fatcat:bxboku2trjh77gb7azfohcfeyu

Tile2Vec: Unsupervised representation learning for spatially distributed data [article]

Neal Jean, Sherrie Wang, Anshul Samar, George Azzari, David Lobell, Stefano Ermon
2018 arXiv   pre-print
To fill this gap, we introduce Tile2Vec, an unsupervised representation learning algorithm that extends the distributional hypothesis from natural language -- words appearing in similar contexts tend to  ...  Our learned representations significantly improve performance in downstream classification tasks and, similar to word vectors, visual analogies can be obtained via simple arithmetic in the latent space  ...  The previous state-of-the-art result used a transfer learning approach in which a CNN is trained to predict nighttime lights (a proxy for poverty) from daytime satellite images -the features from this  ... 
arXiv:1805.02855v2 fatcat:dspizqkalnbbznfpmxirr7buh4

A Survey on Deep Learning-Driven Remote Sensing Image Scene Understanding: Scene Classification, Scene Retrieval and Scene-Guided Object Detection

Yating Gu, Yantian Wang, Yansheng Li
2019 Applied Sciences  
RSISU includes the following sub-tasks: remote sensing image scene classification, remote sensing image scene retrieval, and scene-driven remote sensing image object detection.  ...  As a fundamental and important task in remote sensing, remote sensing image scene understanding (RSISU) has attracted tremendous research interest in recent years.  ...  They performed transfer learning for RSISR by taking into consideration features from the FC and convolutional layers from a wider range of CNN.  ... 
doi:10.3390/app9102110 fatcat:oj3acgbmwnhzppxvvjbsn5cfzq

Quantifying Seagrass Distribution in Coastal Water with Deep Learning Models

Daniel Perez, Kazi Islam, Victoria Hill, Richard Zimmerman, Blake Schaeffer, Yuzhong Shen, Jiang Li
2020 Remote Sensing  
We also demonstrate that using transfer learning techniques for the quantification of seagrass significantly improved the results as compared to directly applying the deep models to new locations.  ...  In addition, we propose a transfer learning approach to transfer knowledge in the trained deep models at one location to perform seagrass quantification at a different location.  ...  For this reason, transfer learning is a common tool used in remote sensing applications. For instance, Hu et al.  ... 
doi:10.3390/rs12101581 fatcat:ffkpacjn4bg6jeeg6u2qvwnpf4
« Previous Showing results 1 — 15 out of 12,257 results